检索结果-内蒙古大学图书馆

Artificial vision system based on porosity heterojunction for in-sensor processing and dynamic trace extraction

DEVICE 2025年第3期3卷

作者： Chang, Ke Yu, Xinna Ren, Wei He, Yunyang Gan, Zhikai Ma, Saiqun Huang, Huanpei Jin, Xianmin Zhang, Yafei Wang, Hui Shanghai Jiao Tong Univ Sch Phys & Astron State Key Lab Adv Opt Commun Syst & Networks 800 Dongchuan Rd Shanghai 200240 Peoples R China Shanghai Jiao Tong Univ Res Inst Micro Nano Sci & Technol Key Lab Thin Film & Microfabricat Technol Minist Educ 800 Dongchuan Rd Shanghai 200240 Peoples R China Shanghai Jiao Tong Univ Sch Elect Informat & Elect Engn Dept Instrument Sci & Engn 800 Dongchuan Rd Shanghai 200240 Peoples R China Shanghai Jiao Tong Univ Qingyuan Res Inst Sch Elect Informat & Elect Engn 800 Dongchuan RD Shanghai 200240 Peoples R China New York Univ Shanghai Sch Art & Sci Shanghai 200124 Peoples R China Chinese Acad Sci Shanghai Inst Tech Phys Key Lab Infrared Imaging Mat & Detectors 500 Yutian Rd Shanghai 200083 Peoples R China Shanghai Jiao Tong Univ Sch Phys & Astron Key Lab Artificial Struct & Quantum Control Minist Educ 800 Dongchuan Rd Shanghai 200240 Peoples R China Shanghai Jiao Tong Univ Sch Mat Sci & Engn 800 Dongchuan Rd Shanghai 200240 Peoples R China

Semiconductor heterostructures can be used to explore band structures for creating optoelectronic devices. In this work, we demonstrate an all-optical-driven artificial vision system based on a porous silicon heterojunction with in-sensor processing capabilities. The abundant surface states of porous silicon enable light trapping and photoelectric conversion, which allow the sensor to exhibit high sensitivity and detect weak light signals. By utilizing differential optoelectronic signals, the integrated device can perform multiple functions within a single unit, such as in-sensor encoding and trajectory tracking of dynamic optical signals. The neuromorphic vision systems based on this photoelectric retina can perform optoelectronic modulation, trajectory reconstruction, and handwritten text recognition and can emulate functions of the human retina with temporal and spatial resolution. The design strategy here can be used to implement in-sensor processing of dynamic optical signals in machine vision systems.

关键词： porous silicon heterojunction retina-inspired sensor artificial vision system in-sensor processing character recognition

来源：评论

学校读者我要写书评

暂无评论

A 22-nm All-Digital Time-Domain Neural Network Accelerator for Precision in-sensor processing

引用

IEEE TRANSACTIONS ON VERY LARGE SCALE INTEGRATION (VLSI) SYSTEMS 2024年第12期32卷 2220-2231页

作者： Mohey, Ahmed M. Leslin, Jelin Singh, Gaurav Kosunen, Marko Ryynanen, Jussi Andraud, Martin Aalto Univ Dept Elect & Nanoengn Espoo 02150 Finland UCLouvain ICTEAM B-1348 Leuven Belgium

Deep neural network (DNN) accelerators are increasingly integrated into sensing applications, such as wearables and sensor networks, to provide advanced in-sensor processing capabilities. Given wearables' strict size and power requirements, minimizing the area and energy consumption of DNN accelerators is a critical concern. In that regard, computing DNN models in the time domain is a promising architecture, taking advantage of both technology scaling friendliness and efficiency. Yet, time-domain accelerators are typically not fully digital, limiting the full benefits of time-domain computation. In this work, we propose an all-digital time-domain accelerator with a small size and low energy consumption to target precision in-sensor processing like human activity recognition (HAR). The proposed accelerator features a simple and efficient architecture without dependencies on analog nonidealities such as leakage and charge errors. An eight-neuron layer (core computation layer) is implemented in 22-nm FD-SOI technology. The layer occupies 70x70 mu m while supporting multibit inputs (8-bit) and weights (8-bit) with signed accumulation up to 18 bits. The power dissipation of the computation layer is 576 mu W at 0.72-V supply and 500-MHz clock frequency achieving an average area efficiency of 24.74 GOPS/mm(2) (up to 544.22 GOPS/mm(2)), an average energy efficiency of 0.21 TOPS/W (up to 4.63 TOPS/W), and a normalized energy efficiency of 13.46 1b-TOPS/W (up to 296.30 1b-TOPS/W).

关键词： Human activity recognition Neurons Accuracy Time-domain analysis Computational modeling Artificial neural networks Computer architecture Wearable devices Energy efficiency sensors Edge computing human activity recognition (HAR) inertial measurement unit (IMU) in-sensor processing multiply-and-accumulate multiply and accumulate (MAC) neural network accelerator smart sensor interface time-domain signal processing

来源：评论

学校读者我要写书评

暂无评论

Toward All-Digital Time-Domain Neural Network Accelerators for in-sensor processing Applications 9

Toward All-Digital Time-Domain Neural Network Accelerators f...

引用

9th IEEE Nordic Circuits and Systems Conference (NorCAS)

作者： Mohey, Ahmed M. Kosunen, Marko Ryynanen, Jussi Andraud, Martin Aalto Univ Dept Elect & Nanoengn Espoo Finland

ISBN: (纸本)9798350337570

Deep Neural Network (DNN) accelerators are increasingly integrated into sensing applications, such as wearables and sensor networks, to provide advanced in-sensor processing capabilities. Given wearables' strict size and power requirements, minimizing the area and energy consumption of DNN accelerators is a critical concern. In that regard, computing DNN models in the time domain is a promising architecture, taking advantage of both technology scaling friendliness and efficiency. Yet, time-domain accelerators are typically not fully digital, limiting the full benefits of time-domain computation. In this work, we propose a time-domain multiply and accumulate (MAC) circuitry enabling an all-digital with a small size and low energy consumption to target in-sensor processing. The proposed MAC circuitry features a simple and efficient architecture without dependencies on analog non-idealities such as leakage and charge errors. It is implemented in 22nm FD-SOI technology, occupying 35 mu m x 35 mu m while supporting multi-bit inputs (8-bit) and weights (4-bit). The power dissipation is 46.61 mu W at 500MHz, and 20.58 mu W at 200MHz. Combining 32 MAC units achieves an average power efficiency, area efficiency and normalized efficiency of 0.45 TOPS/W and 75GOPS/mm(2), and 14.4 1b-TOPS/W.

关键词： Edge Computing Human Activity Recognition HAR Inertial Measurement Unit IMU in-sensor processing Multiply-and-Accumulate MAC Neural Network Accelerator Smart sensor Interface Time-Domain Signal processing

来源：评论

学校读者我要写书评

暂无评论

A Comparative Study of in-sensor processing vs. Raw Data Transmission using ZigBee, BLE and Wi-Fi for Data Intensive Monitoring Applications 11

A Comparative Study of In-sensor Processing vs. Raw Data Tra...

引用

11th IEEE International Symposium on Wireless Communications Systems (ISWCS)

作者： Shahzad, Khurram Oelmann, Bengt Mid Sweden Univ Dept Elect Design Sundsvall Sweden

ISBN: (纸本)9781479958634

Wireless sensor nodes, as typically realized using IEEE 802.15.4 compatible low-power radio transceivers that offer limited throughput, are generally applicable to low-data rate intermittent monitoring applications. In order to realize high sample rate monitoring applications, it requires either transmitting raw data using a high-throughput radio transceiver or performing computation within the sensor node and then transmitting a small amount of information. In relation to a energy constrained wireless sensing node, a quantitative evaluation of raw data transmission using different short range wireless technologies and in-sensor processing is conducted in this paper. The results, associated with the energy consumption of two data intensive monitoring applications, suggest that in-sensor processing resulting in a small amount of data to be transmitted consumes less energy as compared to that of raw data transmission, even under ideal channel conditions.

关键词： ZigBee Bluetooth low energy BLE IEEE 802.15.4 Wi-Fi in-sensor processing wireless sensor

来源：评论

学校读者我要写书评

暂无评论

A Power-Efficient CMOS Image sensor with in-sensor processing and In-Pixel Gain Calibration 65

A Power-Efficient CMOS Image Sensor with In-Sensor Processin...

引用

IEEE 65th International Midwest Symposium on Circuits and Systems (MWSCAS)

作者： Wu, Ke Guo, Yuekang Jin, Jing Liu, Xiaoming Zhou, Jianjun Shanghai Jiao Tong Univ Dept Micro Nano Elect Ctr Analog RF Integrated Circuits CARFIC Shanghai Peoples R China Shanghai Jiao Tong Univ MoE Key Lab Artificial Intelligence AI Inst Shanghai Peoples R China

ISBN: (数字)9781665402798

ISBN: (纸本)9781665402798

This paper presents a power-efficient CMOS Image sensor (CIS) with in-sensor computation and in-pixel gain calibration. The CIS can output 32x32 image or directly process the image in-sensor with analog domain computation. The output of the CIS is sampled by an array of 8bit Single Slope ADC (SS ADC). This CIS features extra low power consumption with highly configurable parameters. Convolution can be done in-sensor with a power efficiency of 131.7 nJ/frame. To suppress the gain fluctuation, a bulk-modulation technique is introduced. The overall gain variation is reduced by 15% in the worst condition. The whole system is designed and simulated in 180nm standard CMOS process. To verify the soundness of the system, the first layer of a 4bit Quantized Neural Network (QNN) on the MNIST dataset is replaced by the proposed CIS. Result shows that the CIS can work under 120 fps and the accuracy of the system only decreases by 1.7% compared to the software. The overall power consumption is only 15.9 mu W, which is greatly smaller than the digital approach.

关键词： CMOS Image sensor in-sensor processing Bulk Modulation

来源：评论

学校读者我要写书评

暂无评论

A Single-Pixel Event Photoactive Device for Real-Time, In-sensor Spatiotemporal Optical Information processing

引用

ADVANCED MATERIALS 2025年第1期37卷 e2406607页

作者： Kumar, Mohit Park, Hayoung Seo, Hyungtak Ajou Univ Dept Mat Sci & Engn Suwon 16499 South Korea Ajou Univ Dept Energy Syst Res Suwon 16499 South Korea

The increasing demand for energy-efficient, sophisticated optical sensing technologies in various applications, from machine vision to optical communication, highlights the necessity for innovations in spatiotemporal information sensing and processing at a nearly single-pixel scale. Traditional methods, including multi-pixel photodetector arrays and event-based camera systems, often fail to provide rapid, real-time detection and processing of dynamic events within the sensor. This shortcoming is particularly notable in handling high-dimensional spatiotemporal data, where the dependency on sequential data input and external processing tools leads to latency, reduced throughput, and heightened energy consumption, thereby impeding real-time parallel data processing capabilities. Here, a carrier-selective, single-pixel, position-sensitive planar photoactive device that integrates spatiotemporal event sensing with inherent short-term memory capabilities is introduced. The proof-of-concept single-pixel event photoactive device enables in-sensor spatiotemporal parallel optical information processing, efficiently managing multibit (>4 bit) data simultaneously and facilitating ultrafast (approximate to 0.4 mu s) recognition of input patterns with low energy consumption (25 femtojoules). Additionally, by adjusting the operating speed from continuous to pulsed light illumination, the sensor array can detect trajectories and absolute position of events, offering in-sensor optical flow detection. This single-pixel event photodetector marks significant advancement toward developing compact, energy-efficient, ultrafast sensors suitable for a wide range of in sensor-based photonic applications.

关键词： event sensing photoactive device in-sensor processing short-term memory spatiotemporal ultralow energy

来源：评论

学校读者我要写书评

暂无评论

Tunable Bipolar Photothermoelectric Response from Mott Activation for In-sensor Image Preprocessing

引用

ADVANCED MATERIALS 2025年

作者： Li, Bowen Lin, Ning Wang, Zhaowu Chen, Baojie Lan, Changyong Li, Xiaocui Meng, You Wang, Weijun Ding, Mingqi Xie, Pengshan Zhang, Yuxuan Wu, Zenghui Li, Dengji Chen, Fu-Rong Chan, Chi Hou Wang, Zhongrui Ho, Johnny C. City Univ Hong Kong Dept Mat Sci & Engn Hong Kong 999077 Peoples R China City Univ Hong Kong Shenzhen Res Inst Shenzhen 518057 Peoples R China City Univ Hong Kong State Key Lab Terahertz & Millimeter Waves Hong Kong 999077 Peoples R China Univ Hong Kong Dept Elect & Elect Engn Hong Kong 999077 Peoples R China Hebei Univ Technol Sch Sci Tianjin 300401 Peoples R China Nanjing Univ Natl Lab Solid State Microstruct Nanjing 210093 Peoples R China City Univ Hong Kong Dept Elect Engn Hong Kong 999077 Peoples R China Univ Elect Sci & Technol China State Key Lab Elect Thin Films & Integrated Device Chengdu 610054 Peoples R China

In-sensor image preprocessing, a subset of edge computing, offers a solution to mitigate frequent analog-digital conversions and the von Neumann bottleneck in conventional digital hardware. However, an efficient in-sensor device array with large-scale integration capability for high-density and low-power sensory processing is still lacking and highly desirable. This work introduces an adjustable broadband photothermoelectric detector based on a phase-change vanadium dioxide thin-film transistor. This transistor employs a vanadium dioxide/gallium nitride three-terminal structure with a gate-tunable phase transition at the gate-source junctions. This design allows for modulable photothermoelectric responsivities and alteration of the short-circuit photocurrent's polarities. The devices exhibit linear gate dependence for the broadband photoresponse and linear light-intensity dependence for both positive and negative photoresponsivities. The device's energy consumption is as low as 8 pJ per spike, which is one order of magnitude lower than that of previous Mott materials-based in-sensor preprocessing devices. A wafer-scale bipolar phototransistor array has also been fabricated by standard micro-/nano-fabrication techniques, exhibiting excellent stability and endurance (over 5000 cycles). More importantly, an integrated in-sensor convolutional network is successfully designed for simultaneous broadband image classification, medical image denoising, and retinal vessel segmentation, delivering exceptional performance and paving the way for future smart edge sensors.

关键词： bipolar photoresponsivity in-sensor processing phase transition photothermoelectric detector

来源：评论

学校读者我要写书评

暂无评论

Thin-film event-based vision sensors for enhanced multispectral perception beyond human vision

引用

INFOMAT 2025年第0期

作者： Li, Kexin Wang, Xiaoting Wu, Yi Deng, Wenjie Li, Jing Li, Jingjie Zhao, Yuehui Chen, Zhijie Yang, Dezhen Yu, Songlin Zhang, Yongzhe Beijing Univ Technol Sch Informat Sci & Technol Key Lab Optoelect Technol Educ Beijing 100124 Peoples R China North China Res Inst Electro Opt Beijing Peoples R China

Dynamic detection is crucial for intelligent vision systems, enabling applications like autonomous vehicles and advanced surveillance. Event-based sensors, which convert illumination variations into sparse event spikes, are highly effective for dynamic detection with low data redundancy. However, current event-based vision sensors with simplified photosensitive capacitor structures face limitations, particularly in their spectral response, which hinders effective information acquisition in multispectral scenes. Here, we introduce a two-terminal thin-film event-based vision sensor that innovatively integrates an inorganic oxide p-n junction with the pyro-phototronic effect, synergistically combining the photovoltaic and pyroelectric mechanisms. This innovation enables spiking signals with a tenfold increase in responsivity, a dynamic range of 110 dB, and an extended spectral response from ultraviolet (UV) to near-infrared (NIR). With a thin-film sensor array, these spiking signals accurately extract fingerprint edge features even under low-light conditions, benefiting from high sensitivity to minor luminance variations. Additionally, the sensors' broadband spiking response captures richer information, achieving 99.25% accuracy in multispectral dynamic gesture recognition while reducing data processing by over 65%. This approach effectively eliminates redundant data while minimizing information loss, offering a promising alternative to current dynamic perception technologies. image

关键词： event-based vision in-sensor processing motion detection multispectral perception

来源：评论

学校读者我要写书评

暂无评论

Transformative Multifunction Deep Ultraviolet Photodetectors for On-Demand Applications: From Fast Optical Communication to Tunable In-sensor Photocurrent Integration

引用

ACS APPLIED MATERIALS & INTERFACES 2024年第21期16卷 27550-27559页

作者： Kumar, Mohit Park, Hayoung Seo, Hyungtak Ajou Univ Dept Energy Syst Res Suwon 16499 South Korea Ajou Univ Dept Mat Sci & Engn Suwon 16499 South Korea

The strategic utilization of photodetectors' transient response could open new frontiers from free-space optical communication to the emerging field of neuromorphic optoelectronics. Contrarily, while communication requires a fast response, neuromorphic applications benefit from a slow and integrative transient photocurrent. By integrating these functionalities in a single device, this study unveils a photodetector with tunable responses, bridging the gap between optical communication and neuromorphic sensing and creating a versatile platform with on-demand applications. Particularly, a Ga2O3-based photodetector was designed, exhibiting a photocurrent on/off ratio close to 10(4), high responsivity of 0.43 A/W, and detectivity 1.22 x 10(13) Jones under deep ultraviolet illumination (lambda similar to 260 nm). The photodetector demonstrates transient time-dependent on operational voltage, ranging from 10(-4) to 0.2 s. The underlying mechanism is attributed to the voltage-dependent balance between photocarrier generation and defect-related recombination, as revealed by electrostatic force microscopy. Additionally, we have demonstrated potential applications, including digital Morse code interpretation, tunable integration of optical inputs within the sensor, one-time readouts, and effective analog Morse code reading. Furthermore, the effectiveness of input information recognition using analog integration, even with anomalies, was demonstrated. This work establishes a versatile approach for tunable in-sensor optical processing, potentially useful for a wide range of applications, from free-space optical communication to neuromorphic sensing.

关键词： optical communication Morse code integration deep UV photodetector in-sensor processing

来源：评论

学校读者我要写书评

暂无评论

A Reconfigurable Layered-Based Bio-Inspired Smart Image sensor 18

A Reconfigurable Layered-Based Bio-Inspired Smart Image Sens...

引用

18th IEEE-Computer-Society Annual Symposium on VLSI (ISVLSI)

作者： Bhowmik, Pankaj Pantho, Md Jubaer Hossain Saha, Sujan Bobda, Christophe Univ Florida Dept Elect & Comp Engn Gainesville FL 32611 USA

ISBN: (纸本)9781728133911

This paper presents a hardware architecture to extract features from an image using the concepts of bio-inspired computing and a method of converting sequential image processing to parallel computational processing units that can execute on the sensor. These computational units are oriented on vertically integrated hierarchical planes and enabled with a region based Attention Module which separates the Regions of Interest (ROIs) from the image. In each layer, the computational units work in parallel and introduce massive parallelism at the pixel level. At the same time, the design saves dynamic power by dynamically enabling and disabling the computational units which ensure high-performance and high-throughput. Moreover, the units are made reconfigurable to support a wide range of machine vision applications by forming a basic structure that is common to all operations and reconfigurable parts for a specific application. Our simulation result shows the design achieves 4.852x power savings on ROIs while processing at 465 Kfps with 800 MHz clock frequency.

关键词： in-sensor processing Reconfigurability Attention Module Image sensor ASIC

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：